On 16 July 2013 00:16, Paul Moore <p.f.mo...@gmail.com> wrote: > > On 15 July 2013 14:53, Brett Cannon <br...@python.org> wrote: >> >> As long as you make sure that sys.path[0] is actually the script location >> then it will work (other things like .pth files, PYTHONSTARTUP, etc. could >> have changed things before your script started execution). But realize that >> a) in Python 3.3 the scripts location will be ./pip.py, not just pip.py, and >> b) if I get my way all paths will be absolute for __file__, so you will have >> to just associate '' with os.getcwd() and then search for the proper >> directory on sys.path. > > > OK, that pretty much tells me that this is a bad idea. It's never going to > be robust enough to work.
Most of the stuff Brett mentioned there shouldn't be relevant for a directly executed script - doing sys.path.remove(os.path.dirname(os.path.abspath(__file__)) should be pretty robust in any scenario. > I'm amazed actually that there's no way to say > "don't add the script location to sys.path", even as a command line option. > It seems like the sort of thing you'd want to make scripts robust, a bit > like -S and -E. You'd think that, but then you'd look at getpath.c and run away (or write something like PEP 432, as I did) :P Cheers, Nick. -- Nick Coghlan | ncogh...@gmail.com | Brisbane, Australia _______________________________________________ Distutils-SIG maillist - Distutils-SIG@python.org http://mail.python.org/mailman/listinfo/distutils-sig
